The Purpose of this study is to:

1. Refine and validate a computerized bedside decision support tool blood glucose management in critically ill adult and pediatric ICU patients.
2. Monitor how often low blood sugar levels occur during use of the bedside tool.
3. Determine how the computerized tool effects the workload of the ICU nurses.

This first phase will be a prospective cohort study. We will validate the bedside tool in at least 4 adult and at least 4 pediatric ICUs and if necessary we will modify the bedside tool. The computerized bedside tool will be considered refined if >90% of the generated instructions are accepted and if the percent of glucose values in the 70-110 mg/dl range are equivalent to our pilot experience rate of 55%, and if the rate of glucose values less than or equal to 40 mg/dl is less than 0.5% of glucose measurements.

Inclusion Criteria:

1. Pediatric ICU: Proposed indications by pediatric clinicians at participating ICUs for glucose control in their ICUs would include patients who require mechanical ventilation for greater than 24 hours and/or require a vasoactive infusion (e.g. dopamine >3 mg/kg/min, dobutamine, epinephrine, or vasopressin).
2. Adult ICUs: Proposed indications by clinicians in participating ICUs for glucose control in participating adult ICUs include critically ill patients with an anticipated ICU length of stay of 3 or more days.

Exclusion Criteria:

1. Pregnancy (negative pregnancy test required for females of child-bearing age)
2. Age less than one month
3. Inborn errors of metabolism that the clinician suspects will affect glucose homeostasis
4. Acute or chronic liver disease with any documented episode of blood or plasma glucose <60 mg/dl within the 24 hours prior to study entry
5. Diabetic Ketoacidosis (critically ill patients with insulin dependent diabetes not in ketoacidosis will be eligible if attending physicians intend to use intravenous insulin as part of ordinary care)
6. Severe chronic liver disease (Child-Pugh score >10)